Ada Zhang

This study evaluates deep learning models for gene expression prediction in disease contexts, finding they perform poorly in disease-relevant applications despite high overall correlation.

This paper introduces PentaMind, a deep learning model that extracts cognitive features from hand-drawn intersecting pentagons, improving detection of cognitive decline in older adults.

The paper identifies proteins and molecular processes related to brain connectivity by integrating neuroimaging, proteomic, and gene expression data from human brain regions.